We show how to find regions of interest (ROIs) in trajectory databases. ROIs are regions where a large number of moving objects remain for at least a given time interval. Previous techniques use somewhat restrictive definitions for ROIs, and are parameter-dependent. They require sequential scanning of the entire dataset to find ROIs when the ROI parameters change. Our approach is parameter independent, so that the user can quickly identify ROIs under different parametric definitions without rescanning the whole database. We also generalize ROIs to be regions of arbitrary shape of some predefined density. We have tested our methods with large real and synthetic datasets to test the scalability and verify the output of our methods. Our methods give meaningful output and scale very well.